seeks the lowest levels attainable, these currents will find a place at the bottom of the ice-sheet, and wear away the ground-moraines and other débris. Hence banks of earth or clay will be found continuous from the ice-cliff to as great a distance toward deep water as the currents have power to transport the material. Marine animals live in and upon these banks, and leave their remains in them. Hence the deposit is analogous to that series of marine clays called Champlain by me in 1861, and occurring so plentifully up to three hundred feet in the St. Lawrence Valley, and to one hundred and sixty feet along the coast of northern New England and the Provinces. It is obvious that they were contemporaneous with the glacial moraines upon the land, and not entirely consecutive, as so many have supposed. This was the position assumed by the late Professor Agassiz, and confirmed by our own published observations in the "Geology of New Hampshire," describing the occurrence of fossiliferous beds between the lower and upper till of Portland, Maine.
The Eastern American Ice Area.—The latest generalizations indicate that some part of the Labrador Peninsula may be considered as the center from which the ice has radiated over the Dominion of Canada and the northern United States east of the Rocky Mountains. By regarding the Greenland and polar areas as independent of the Labrador sheet, though possibly confluent at the time of maximum glaciation, a multitude of difficulties are removed, and our American glaciers are seen to have been subjected to the same laws as the several ice-fields of other parts of the world. This area extends from Baffin's Bay to Dakota, and from Hudson's Bay, or the Arctic Archipelago, to a line drawn from the Great Banks of Newfoundland through New Jersey, southern Ohio, etc., to the plains east of the Rocky Mountains.
Most of this territory exhibits a southwesterly course of glaciation. This is well shown over the highlands between Hudson's Bay and the St. Lawrence Valley, the valley itself, western New York, Ohio, Iowa, Minnesota, Manitoba, and so on to the extreme western limits. It is very prominent from Lake Superior, near the international boundary, westward to the Rocky Mountains. In eastern New York and the Champlain and Hudson Valleys the course is southerly. In New England the dominant direction is southeasterly, and the same is true of New Brunswick and Nova Scotia. The exceptional deviations are due to local influences, exerted in the decline of the age. The facts in hand for Newfoundland are too few for a satisfactory delineation, while not in disagreement with these generalizations. On the east coast of Labrador there arc several fiords, as if there had been an ice sheet upon the upper part of the peninsula, moving northeast and east. Professor O. M. Lieber's sketches in the Coast Survey report suggest a local glaciation, and not such a general smoothing as would be manifested in cast' the ice had come from the opposite Greenland coast.